Editor's Note: At the 2024 SABCS conference, Professor Charles Geyer from the University of Pittsburgh, USA, presented the latest progress of the NSABP B-59/GBG-96-GeparDouze study. The study investigated whether the combination of atezolizumab and neoadjuvant chemotherapy could improve event-free survival (EFS) in patients with triple-negative breast cancer (TNBC). Although the study failed to meet its primary endpoint, showing no significant statistical improvement in EFS, subgroup analyses indicated that some patients appeared to benefit from the treatment. Professor Charles Geyer was interviewed on-site at SABCS by Oncology Outlook. In the interview, he stated that despite the negative results, the study provided valuable data for the treatment of patients with triple-negative breast cancer.Oncology Frontier: Please share the key findings of the NSABP B-59/GBG-96-GeparDouze study. In particular, what are the highlights regarding the efficacy and safety of atezolizumab when added to neoadjuvant chemotherapy and then used as adjuvant therapy in patients with stage II/III triple-negative breast cancer (TNBC)?
Professor Charles Geyer:I had privilege today to present results of a phase III study in early triple negative breast cancer that my colleagues in the NSABP and German Breast Group had been working on together for quite a few years now, probably about eight years. The study was a double blind phase III trial for women who had stage II and III triple negative breast cancer at presentation and were candidates for curative neoadjuvant chemotherapy with our standard programs that we use here in the US of taxane carboplatin followed by anthrocycline cyclophosphamide. In our country now and many countries around the world, of course, the standard has become to co-administer the checkpoint inhibitor pembrolizumab（Pembro） during neoadjuvant therapy as well as adjuvant therapy based on KEYNOTE-522.
Our study actually started about eight months after KEYNOTE-522 started, and both studies were interested in the same question. Could a checkpoint inhibitor added to chemotherapy improve path CR, improve overall survival, and hopefully improve survival.
The Pembrolizumab study was an ambitious global study that was sponsored by the company, had multiple sites, patients around the world. Our study was a smaller study limited to Europe and U.S., largely Germany and Spain, and actually in Europe. So it took us twice as long to accrue our study. And that’s why we learned the results of the KEYNOTE-522 several years before we were able to report our study. So our study actually started as a contemporary trial, but we just now had sufficient events to report our primary endpoint of event-free survival. And so, you know, it was just a question: could an inhibitor of PD-L1 as opposed to an inhibitor of PD-1, which is what Pembro inhibits, what kind of effect could it have in triple negative breast cancer.
Our study had some differences that I think people will be interested in. The KEYNOTE-522 study had a 2:1 randomization with about a thousand patients. We had 1:1 randomization with 1550 patients, so our trial was larger.
We both had similar criteria for being eligible for the study in terms of basically T2, T3, N0 or no positive patients. So we did recruit different population cohorts though because the KEYNOTE-522 study was a global study, had a broad representation. For instance, a substantial percentage of Asian patients were in KEYNOTE-522. In our study, we had relatively few Asian patients. Our studies were largely white patients just because that’s the makeup of Germany and Spain by and large. So our study differed in terms of the patient population. Half of their patients were no positive, only 40% of our patients were no positive. We also decided to allow dose -dense chemotherapy to be given. The KEYNOTE-522 study had specified every three weeks. We felt like, well, the standard is dose-dense, at least we should let doctors and patients decide. So these are some of the differences between the studies. We also, because of the impact of the CREATE -X study while we were recruiting our patients, had to ultimately allow patients who did not have a pCR to get Capecitabine.
And so half of our non-pCR patients got Capecitabine. So there are differences in the patient populations that came into the study. There’s also differences in the therapy that I think will be of interest, you know, doing the cross trial comparisons. Bottom line to our study was that we did not meet our primary endpoint. We did not show that adding atezolizumab to standard chemotherapy resulted in a statistically significant improvement in event-free survival. So we have to consider our study negative for the goal of showing clear superiority. However, when we look at our subsets, we did see our node -positive patients seem to benefit.
Patients with larger tumors seem to benefit. Patients who had high-TIL seem to benefit. So we have activity in our trial. It just wasn’t enough across the whole patient population for us to conclude the trial was positive. That said, we have collected specimens from all the patients at baseline. We had on-therapy biopsies in 500. We have the non-pCR patients so we have a real great opportunity to do what we hope will be very informative translational research to try to understand can we maybe identify some predictors of patients more likely to benefit because that’s still a need in triple negative breast cancer. We treat everybody. We haven’t been able to find biomarkers that tell us who really benefits a lot and who doesn’t benefit at all. We just have to treat everybody because we don’t have biomarkers. We’re hopeful that our study will be able to contribute to that critical question of biomarkers.
Oncology Frontier: What were the key inclusion and exclusion criteria during patient enrollment? In addition, could you explain in detail how patients were stratified according to region, tumor size, chemotherapy regimen (AC/EC), and lymph node status?
Professor Charles Geyer:As far as eligibility went for the study, we did require central testing of a diagnostic core biopsy so we could confirm the patient had triple negative breast cancer. Tumors had to be ≥2cm if a patient was no negative. If they were no positive, they could be down to one centimeter of any size. We did not include T4 tumors, so it was that intermediate risk group. Patients basically had to have normal cardiac, renal, hepatic function to be on a trial. ECOG status of zero to one. The exclusion criteria were largely things that are contraindications to using checkpoint inhibitors.
So it was a pretty broad eligibility criteria. The median age of our patients was 49, so it was a younger patient population, but a pretty healthy population. In terms of our stratification variables that we chose for our study, since we were doing recruitment here in the U.S. and Canada. The Germans were going to be doing recruitment in Spain and Germany. So we made continent, so to speak, Europe or U.S., one of our stratification factors. We made clinical nodal status, a stratification factor, positive versus negative. We also made tumor size. We decided to split our three centimeters. The typical T and M is zero to two, two to five, five or greater. But we knew that if we kind of wanted to have roughly equal groups, so we said up to three and above three, that was a stratification variable. We included PD-L1 as a stratification, PD-L1 staining using the SP142 assay as one of our stratification variables. And then the final stratification variable was the regimen, the rate that they gave the AC or EC, dose down, was it every two weeks? Every three weeks. That was the other stratification variable that we used.
Oncology Frontier: Based on the results of the NSABP B-59/GBG-96-GeparDouze trial, what do you think are the future research directions worth exploring in the treatment of TNBC? What are your expectations and suggestions for further improving the treatment effect and quality of life of TNBC patients?
Professor Charles Geyer:Yeah, I mean, I think for us what we’re hopeful we can get out of our study, since we didn’t get there with an efficacy indication. A big need that many of us still feel is that it would be really helpful to be able to identify the subset of patients who benefit from receiving the checkpoint inhibitor. We used to give chemotherapy to all women with negative lymph nodes, knowing that we probably were not helping many women, but we’re helping some, a few, a lot. And we just didn’t have, we couldn’t do the tools. So now we have genomic markers like the Oncotype DX Recurrence Score, Mammaprint, other things that help us know which ER -positive patients need chemotherapy. That was a huge advance.
We probably have the same thing here in triple negative breast cancer. We know they need chemotherapy. When we added a checkpoint inhibitor, we do better overall. So far, there’s not been anything identified to tell us who does and doesn’t benefit other than having really small tumors. So if we can find some kind of a predictive biomarker, I think that would be very, very helpful to tell us who really needs it and doesn’t need it. Another thing that I think will move it forward is seeing how the ADCs come in, Dato -DXd, others, there’s a lot of them. If they can be even more effective than Capecitabine for the non-pCR, we’re also pulling them forward into the neoadjuvant setting. The potential to fully realize the ADCs is probably the next step, but there’s also then trying to further augment the immune response, but I think that’ll kind of be the, not the next thing, it’ll be the following thing in my view.
